Cristian-Dragos Obreja

The concept of energy analysis is defined and applied to industrial processes. The study discusses the significance of choosing the definition of efficiency, system limitations and problem definition. The purpose of this report is to show the simplicity and value of using the concept of energy analysis in the analysis of industrial processes and to develop conventions and standards in the field of process efficiency. In the situation of the Republic of Moldova, the implementation of projects to improve energy efficiency and the use of renewable energy sources contributes to reducing the degree of the energy dependence of the country and also represents a national contribution in the fight against climate change.

The main goal of this study is to create an innovative approach to retinal vessel detection using the gradient pyramid fusion algorithm to improve edge continuity and measurement precision in retinal images. Traditional methods, like wavelet transform and guided filter, face challenges with background noise, artifacts and uneven illumination, which can distort vessel measurement. The proposed fusion method manages to overcome these limitations by combining the strengths of traditional techniques, which creates more continuous edges through gradient fusion across multiple scales, thus managing to also limit the number of image artifacts. We used images from the DRIVE database, to evaluate the fusion algorithm’s precision, with results showing improved vascular tree detection and continuous edges, a reduction in the number of artifacts and improved measurement accuracy. The results show that this image fusion method improves the retinal image analysis, thus helping in early disease diagnosis.

The world’s rapidly growing demand for raw manganese has made it increasingly important to develop methods for the economic recovery of manganese from secondary sources. The current study aims to present possible ways for the recycling and reuse of silico-manganese slag landfilled in Tulcea, City on the Danube River close to the Danube Delta Biosphere Reserve in order to save the natural resources raw of manganese. In the last three decades, the ferroalloy production plant has over 2.6 million tons of slag. Slag dumping constitutes a significant source of air, water and soil pollution, which adversely affects the environment and human health. Mn present in the slag dump is an environmental pollutant with potentially toxic effects. The results obtained with a leaching method to recover manganese from slag shows two efficient ways to valorize manganese from solid fraction (54%) with size particles between 80 and 315 µm and/or reuse the leaching medium (56% Mn) with a slag size of <80 µm. The motivation of our research is the possibility to recover manganese from slag by saving natural resources of raw of manganese and the remaining fraction can be used as aggregate sources (construction and road rehabilitation by saving extract mineral aggregates and agriculture), in order to decommission the slag dump. The proposed research is in concordance with the sustainable use of natural resources for the achievement of sustainable development of the 2030 Agenda and Waste Management Legislation due of the huge ecological costs regarding non-conforming waste dumping. If we consider the cost-benefit analysis, the environmental future is more important the human health and the benefits on the quality of the population’s health and the environment which are not non-measurable in monetary value.

In the automatic detection framework, there have been many attempts to develop models for real-time melanoma detection. To effectively discriminate benign and malign skin lesions, this work investigates sixty different architectures of the Feedforward Back Propagation Network (FFBPN), based on shape asymmetry for an optimal structural design that includes both the hidden neuron number and the input data selection. The reason for the choice of shape asymmetry was based on the 5–10% disagreement between dermatologists regarding the efficacy of asymmetry in the diagnosis of malignant melanoma. Asymmetry is quantified based on lesion shape (contour), moment of inertia of the lesion shape and histograms. The FFBPN has a high architecture flexibility, which indicates it as a favorable tool to avoid the over-parameterization of the ANN and, equally, to discard those redundant input datasets that usually result in poor test performance. The FFBPN was tested on four public image datasets containing melanoma, dysplastic nevus and nevus images. Experimental results on multiple benchmark data sets demonstrate that asymmetry A2 is a meaningful feature for skin lesion classification, and FFBPN with 16 neurons in the hidden layer can model the data without compromising prediction accuracy.